P91 welded pipe 1

[carletes]

Hello all! We are condidering the possibility of using welded alloy pipe A691 P91 Cl. 42 in a hot reheat system (30 bar, 575ºC) instead of semaless A335 P91 pipe. Manufacturer tell us that all its properties after normalized and tempering are the same than seamless pipe but we are not very sure. We have heard some things about creep properties of Heat Affected Zone that don't sound very well. Any comment or recommendation?

Best regards,

 

[athomas236]

4. Hot bending

(a) The problem

As previously mentioned P91 depends on a specific microstructure for its elevated creep strength properties. Any thermal process such as inter-critical heat treatment or tempering that changes the microstructure could substantially reduce the creep strength so that it approaches, in the worst case, that of 2 ¼%Cr, 1%Mo.

(Note: Inter-critical heat treatment is heating of the material into the temperature range where the microstructure begins to transform from ferrite to austenite. During heating, this process of transformations starts at the lower critical temperature and is completed by the upper critical temperature.)


(b) The solution

Recommendations are before the Code committees to ensure that if it is necessary to normalise and temper P91 material then this process is carried out in the most appropriate temperature ranges.

The temperature ranges recommended for normalising and tempering are 1040-1080°C and 730-800°C respectively.

There is a recommendation in the ASME paper that if any pipe component is heated “locally” above 800°C then the component should be re-normalised and tempered in a manner that will eliminate the zones that have been heated into the inter-critical range.

(c) Other issues

ASME B31.1, clause 129.3.1 requires that hot bending/forming should be carried out at a temperature above the lower critical temperature minus 56°C, i.e. approximately 754°C.

The ASME paper says that significant reductions in creep strength can occur if P91 is exposed to temperatures of the lower critical temperature minus 28°C, i.e. approximately 782°C.

Therefore, to comply with B31.1 and prevent reductions in creep strength while avoiding the need to re-normalise and temper means hot bending should be carried out in the temperature ranges 754-782°C. If this temperature range were selected for hot bending then practical difficulties are expected in maintaining the material temperatures within this range. There is the complication that the lower critical temperature is dependent upon material composition which will vary from batch to batch.

It is understood that hot bending is usually carried in the range 750-950°C.

B31.1, clause 129.3.3(A) states only that any post-hot bending/forming heat treatment should be specified by the designer. There have been cases in the USA when bends that were formed at 1090°C have not been re-normalised and tempered.

If view of the above, we should require all hot bends should be re-normalised and tempered and that the necessary records should be provided to confirm that this has been carried out. This agrees with the requirements stated in the T91/P91 Book.

In the case of ??? project, we should request details of hot bending procedures together with details of any subsequent heat treatment.

 

[athomas236]

5. Post weld heat treatment (PWHT)

(a) The problem

Research has shown that the upper and lower critical temperatures (together with the start and finish temperatures for martensite formation during cooling) are significantly affected by the manganese and nickel contents. The higher the manganese and nickel contents the lower these temperatures.

P91 material has a specified maximum nickel content of 0.4% whereas the applicable AWS specification for weld filler metal has nickel content up to 1%.

The concern is that the high nickel content of the filler metal in combination with the effect on the critical temperatures could result in the incorrect microstructure being present in the welds. The effects of this are expected to be a reduction in the long term creep strength.

(b) The solution

Recommendations are before the Code committees to change the requirements for PWHT.

The PWHT requirements before Code committees are:


%(Mu + Ni) %Ni Max Temp C

?1.5 - 774
<1.5 but ?1.0 >0.4 774
<1.5 but ?1.0 ?0.4 788
<1.0 ?0.4 800
<1.0 >0.4 788

The PWHT requirements in B31.1 are:

Holding temperature 700-760C

Holding time
t ? 50 mm 2 hours

t > 50 mm 2 hours plus 15 min
per 25 mm over 50 mm

It is evident that the B31.1 PWHT requirements and those in the ??? project weld procedure are completely different to those being considered by the Code committees. This is a matter that clearly needs to be resolved with the contractor.

(c) Other issues

In the ASME paper it is stated that requirements for cooling below the preheat temperature before PWHT will be considered when definitive data is available.

B31.1 says that after commencing welding, preheat temperature shall be maintained until PWHT is carried out. This requirement of B31.1 is not mentioned in the Umm Al Nar weld procedure.

There is evidence that after welding P91 should be cooled to 100°C before PWHT to achieve the required microstructure. In Table 9.14.1-1 of the British/European piping code BS/EN 13480-Part 4, states that welds in P91 materials should be cooled before PWHT to produce transformation to martensite. There are, however, no details of how this should be done in terms of temperatures or cooling rates.

The recommendation to cool to 100°C before PWHT is also included in the T91/P91 Book.

We should require all welds to be cooled in accordance with BS EN 13480-Part 4. In addition, the necessary records should be provided to confirm that this has been carried out.

For ??? project, this is a matter that needs to be resolved with the contractor because the weld procedure does not include any requirement for cooling or the maintenance of preheat temperatures prior to PWHT.

 

[athomas236]

6. Cold bending

(a) The problem

Any cold working such as cold bending affects the microstructure of the material. The greater the cold working, the greater the effect on the microstructure. There is a level of cold working that will cause significant reductions in the materials rupture strength.

(b) The solution

ASME is considering placing a limit on the extent of cold working above which it will be necessary to re-normalise and temper.

Based on available data the likely limit for P91 is 15-20% (or less).

B31.1, clause 129.3.3, requires all cold formed bends to be heat treated (in accordance with requirements for PWHT) if the pipe is 4 inch or larger or has a thickness of 0.5 inch or greater.

The British/European piping code BS EN 13480-Part 4 requires heat treatment as shown below.

Mean bend radius Pipe OD Heat treatment
? 1.3 OD All Required
> 1.3 OD but < 2.5 OD ? 142 mm Not required
> 1.3 OD but < 2.5 OD > 142 mm Required
? 2.5 OD All Not required

If view of the above, we should require all cold bends should be re-normalised and tempered in accordance with BS EN 13480-Part 4 except that 142 mm should be replaced with 100 mm. In addition, the necessary records should be provided to confirm that this has been carried out.

In the case of ??? project, we should request details of cold bend procedures together with details of the subsequent heat treatment

 

[athomas236]

7. Stress corrosion cracking

(a) The problem

There is evidence that welds in P91 material may be susceptible to stress corrosion cracking.

(b) The solution

The ASME paper recommends that when P91 has been welded the components should be kept completely dry until PWHT is carried out.

The T91/P91 Book recommends that the maximum period between welding and PWHT should be one week and that during this period the components should be kept dry.

The weld procedure for ??? project does not mention this requirement which should be clarified with the contractor.

 

[davefitz]

Don't forget : the rate of cooling following normalization at 1900 F must be faster than -9F/min in order to avoid formation of ferrite. See the TTT curves for this alloy.

 

[athomas236]

Thanks

 

[athomas236]

Well we had a response from the contractor today. He says all piping is completely and wholey in accordance with B31.1.

If we want to consider other requirements such as presented at the 2004 ASME P&PV conference or the T91 Vallourec and Mannesmann Book we should issue a variation to the contract.

Promises to be an interesting time over the next few months.

athomas236

 

[MEMO4EVER]

what is the ideal fabrication & Erection Methode for A335-P91 Pipes (High alloy steel pipes)?
And how could i weld these pipes?